Author Topic: IGBT failure in DRSSTC  (Read 1290 times)

Offline Felix Bieri

  • High Voltage Enthusiast
  • *
  • Posts: 6
  • Karma: +0/-0
    • View Profile
IGBT failure in DRSSTC
« on: September 20, 2018, 02:42:09 PM »
 [ Invalid Attachment ]  [ Invalid Attachment ] Greetings

I have a problem regarding my DRSSTC. My IGBT's die every time i power up my gate driver. As i have already burnet up 4 of my IGBT modules and the prize for this project now slightly climbs to a costs in which i can't really afford it with my students budget, I'm finally here to ask for help in this more then competent forum (thanks for all the answers i found in threads for my the project so far). I have currently dismantled my DRSSTC for redesign and upgrade purpose and some parts are still on the way from china so i can rebuild it.

The stats of the coil right now is:
Primary coil: 14.76uH
Primary cap: 200nF 6kV
Secondary coil: 116.84mH
Secondary cap: 31.62pF
Resonant Freq.: 65.49kHz
 
IGBT (SKM200GB128D): 1200V 300A continuous (1200A peak)
2 440V TVS over every IGBT
2 33V Zener over ervery IGBT gate
1 Snubber cap: 5uF 800V
Alu-Supercaps : 9mF 800V
 
Driver: UD2.7C from loneoceans
Supply Voltage: 565V (I can power it up with 325V aswell and i have tested everything with 24V before i went up to 565V.... maybe I should have waited at 325 first......)
My GDT consist of 24:24 winding around a ferrit.
(See attachments for pictures)

The improvements would be:
1x 20uF Snubber over each IGBT module
Redesign of the bridge for maintainability purpose
Primary cap: 400nF 12kV

Also, is there a way to calculate my primary peak current? For many calculations the current is needed and some calculations i have found don't make sense to me. Do i just assume a current and implement my OCD for a bit less than my IGBT can switch (1200A peak)? This doesn't seem appropriate and i would like to have propper numbers before i burn some more IGBT's  :o.

I'm very grateful for any suggestions regarding my Problem.
Greetings
Metric system ftw :)

Offline Teravolt

  • High Voltage Technician
  • ***
  • Posts: 100
  • Karma: +2/-0
    • View Profile
Re: IGBT failure in DRSSTC
« Reply #1 on: September 20, 2018, 09:10:10 PM »
what is your gate drive waveform look like?
how much drive voltage is at the gate?
is your gate drive waveform messy or wierd or clean?
do you have enough dead time?
what does the wave form at the tesla look like?

  I think you may have shoot through or the gates are being damage or your feedback is bad in some way. before you power every thing up check all of your semiconductors including your bridge recifire and drive electonics. after a failure things can be weakend or damaged. make shure that the bridge changes state at the same time that the tesla is oscilating so your zero cross is correct and there is no phasing problems. if that happened your igbt diode may be comutating or you have hard switching going on.

Offline Mads Barnkob

  • Administrator
  • High Voltage Expert
  • *****
  • Posts: 1020
  • Karma: +16/-0
  • Denmark
    • View Profile
    • Kaizer Power Electronics
Re: IGBT failure in DRSSTC
« Reply #2 on: September 20, 2018, 09:30:33 PM »
Hi Felix and welcome to HVF :)

I think you have a sound design here and there is no obvious faults jumping right at me, so I would start looking at the driver/GDT and see how they are driving the IGBTs. SKM200 does not just die like that, but if you start out on full DC bus voltage, you will never know if something is wrong before it just goes bang.

Do they die silently or with a bang?

You need to test with a lower voltage on the DC bus and observe the current use, if it rises dramatically above the forward voltage, you most likely have a short circuit of the IGBTs directly across the DC bus. Either due to driving them at wrong phasing, bridge construction fault or some failed part.

Remember to change everything after the GDT when you have had a IGBT failure! Gate resistors, TVS and diodes might look fine, but could be shorted or damaged! Solder it apart and check each component on its own.

Here is a good resource for trouble shooting the GDT waveform: http://www.richieburnett.co.uk/temp/gdt/gdt2.html
http://www.kaizerpowerelectronics.dk - Tesla coils, high voltage, pulse power, audio and general electronics

Offline Felix Bieri

  • High Voltage Enthusiast
  • *
  • Posts: 6
  • Karma: +0/-0
    • View Profile
Re: IGBT failure in DRSSTC
« Reply #3 on: September 20, 2018, 09:45:44 PM »
Thanks a lot for the fast reply. And I'm glad I haven't really designed something completely wrong. If I remember correctly I had steep switching edges and no ringing. But I may have had overshots at my GDT... I can't recall caus my memory stick with the pictures on it is lost in space  :'( and as of now i can't test it again because my Driver broke too from a broken Power Supply.... I really need my stuff to get fixed....  ::) :o
To Mads question. Yes they die with a big bang :)
Anyway thanks a lot. Now i have some clues of what to look at when I'm off testing it again once it's rebuilded.

I'll be updating this thread then as soon as I have new data or questions. If anyone has further suggestions I'd gladly hear them.
Thanks a lot again!
« Last Edit: September 20, 2018, 11:32:12 PM by Felix Bieri »
Metric system ftw :)

Offline Hydron

  • Administrator
  • High Voltage Engineer
  • *****
  • Posts: 231
  • Karma: +6/-0
    • View Profile
Re: IGBT failure in DRSSTC
« Reply #4 on: September 21, 2018, 04:52:34 PM »
As Mads said, there isn't anything obviously wrong, so more info would help. Photos/diagrams, scope screenshots of gate drive etc.

Remember that you can hook up a coil to a bench supply (or a few small supplies in series) for testing at say ~50-100V - even if they would current limit if run continuously you'll still be able to capture the current/voltage waveforms from the bridge for a single pulse. If that is successful then I'd try and ramp it up slowly from there - if you don't have a variac even stopping at rectified mains (rather than doubled mains) would be better than going all the way in one jump.

Offline Felix Bieri

  • High Voltage Enthusiast
  • *
  • Posts: 6
  • Karma: +0/-0
    • View Profile
Re: IGBT failure in DRSSTC
« Reply #5 on: October 02, 2018, 03:32:26 PM »
My snubbers (DHA1200K20#) have finally arrived: http://www.dawn-cap.cn/upload/201609/DAWNCAP-DHA%202015.pdf

I have yet to receive all my other parts tho. So here a fiew questions first for further planning.
Im gonna replace my primary capacitors with DTR6000K0.40-*-# also from dawncap: http://www.dawn-cap.cn/upload/201806/DAWNCAP-DTR%202015.pdf
As stated in the datasheet the Ipeak these capacitors can bear is 1400A, which is confirmed when i calculate the Ipeak with: Ipeak = C * dV/dT. As stated on Mads site: http://kaizerpowerelectronics.dk/theory/good-mmc-capacitors/. So i am not allowed, for the sake of my dear capacitors, to go with my OCD over this value of 1400A? Is this correct? Or is the OCD limited with the peak current value of the IGBT aka. 1200A? I'm kind of confused at what level I should set my OCD so i can prevent too strong Currents. :o

Edit: My confusion is probably caused because I've seen some people talk about 2000A peak currents in the primary with CM600DU-24FA IGBT. In the datasheet the Ipeak is set to 1200A tho..... how can they have so high peak currents without breaking the IGBT?
« Last Edit: October 02, 2018, 08:45:07 PM by Felix Bieri »
Metric system ftw :)

Offline Mads Barnkob

  • Administrator
  • High Voltage Expert
  • *****
  • Posts: 1020
  • Karma: +16/-0
  • Denmark
    • View Profile
    • Kaizer Power Electronics
Re: IGBT failure in DRSSTC
« Reply #6 on: October 03, 2018, 10:44:41 AM »
I am looking forward to hear about your results and experiences with the Dawncaps, I have also looked at them before and on paper they do look good to the price, but I never tried them.

Maximum peak current all comes down to dissipated heat, so with more cooling, larger thermal mass and such we can overload the components but at the cost of their expected life time. To calculate the maximum possible switching speeds and currents for a IGBT, check out my guide on that. It is simply a matter of soft switching being less stressful on the IGBTs and that we aim for a certain die temperature rise per pulse: http://kaizerpowerelectronics.dk/tesla-coils/drsstc-design-guide/igbts/

So if you push higher peak currents through something, it just means higher temperature and if you keep adding that up, it usually ends with fire ;) 
http://www.kaizerpowerelectronics.dk - Tesla coils, high voltage, pulse power, audio and general electronics

Offline Felix Bieri

  • High Voltage Enthusiast
  • *
  • Posts: 6
  • Karma: +0/-0
    • View Profile
Re: IGBT failure in DRSSTC
« Reply #7 on: October 16, 2018, 11:54:48 AM »
@Mads: Im quite sure i found an error on your page. Chapter "Thermal impedance": You compare the Thermal impedance of 300BPS with a dutycycle of around 0.06 with the thermal impedance of the 0.5 Dutycylce given in the datasheet. You get a factor 2.66 which is wrong i think. 0.17K/W / 0.015K/W (@ D=0.06 from shematic) is around factor 11.33 lower not 2.66. I think you compared Thermal impedance of the 0.5 dutycycle with the dutycycle of 300BPS directly.

Also i have finished repairing my Driver, GDT and Interrupter (I currently use an interrupter module to convert MIDI from a Keayboard with maxFreq = 554.37Hz, a Cis5)

I tuned the Interrupter to always have a Dutycycle of 10% at the maxFreq. So for this frequency it would be around 200us. The feedback of 65kHz was simulated with an oszillator. Images of the tests and results are attached. Are these values plausible? (Its an ancient oszilloscope... pls have mercy about the quality  ;D)

I just need my primary caps now so i can rebuild the rest of my coil and test it further.

Greetings
Metric system ftw :)

Offline Mads Barnkob

  • Administrator
  • High Voltage Expert
  • *****
  • Posts: 1020
  • Karma: +16/-0
  • Denmark
    • View Profile
    • Kaizer Power Electronics
Re: IGBT failure in DRSSTC
« Reply #8 on: October 18, 2018, 08:02:24 PM »
Thanks for catching that error Felix! I corrected the page and added a little more text to make it less confusing.

Your oscilloscope shots looks fine and corresponds to your design criteria, 10% duty cycle is a good maximum for MIDI, just watch the temperature if you have many long notes
http://www.kaizerpowerelectronics.dk - Tesla coils, high voltage, pulse power, audio and general electronics

Offline Felix Bieri

  • High Voltage Enthusiast
  • *
  • Posts: 6
  • Karma: +0/-0
    • View Profile
Re: IGBT failure in DRSSTC
« Reply #9 on: January 08, 2019, 05:33:55 PM »
Hello again

I've been busy over the holidays and still am because of exams :S. But still I managed to finish constructing my DRSSTC for now. I managed to test some parts as well. As of now I've verified that the DC Voltage works perfectly fine (for testing purpose I drive with a onephase variac with 50VAC). My driver works together with my interrupter just fine. Also my switching of the resistors works as intended -> I switch a resistor in series for loading the supercaps at the beginning to limit current, after 40sec i shortcircuit the resistor. Because I want to drive the coil at 400V in the end and no threephase variac available this is my solution.

The issue i face now is my IGBT's are now not really switching and my driver doesn't really "start" to output signals. I probably did not fine tune my driver correctly for the now added inductance of the 15:15 transformer for the gate driver. Maybe I have to delay more tests forther caus of exams but never the less here some pictures of the coil as of now.

Greetings ;D
Metric system ftw :)

High Voltage Forum

Re: IGBT failure in DRSSTC
« Reply #9 on: January 08, 2019, 05:33:55 PM »

 


* Recent Topics and Posts

post Re: SSTC interruptor not working
[Solid state Tesla coils]
Mads Barnkob
Today at 11:46:44 AM
post Re: Welcome new members, come say hello and tell a little about yourself :)
[General chatting]
Mads Barnkob
Today at 11:45:11 AM
post Re: "LightningStorm" VTTC project
[Vacuum tube Tesla coils]
Mads Barnkob
Today at 11:40:18 AM
post Re: My first DRSSTC on bricks
[Dual Resonant Solid State Tesla coils]
Netzpfuscher
Today at 07:46:25 AM
post Re: Si-wafers, ideas?
[Electronic circuits]
Weston
Today at 07:07:57 AM
post Re: Help for people buying the "12-48 Volt 1800/2500 Watt ZVS induction Heater"
[Electronic circuits]
hightemp1
Today at 05:45:55 AM
post Re: My first DRSSTC on bricks
[Dual Resonant Solid State Tesla coils]
Laci
June 24, 2019, 09:26:47 PM
post Re: Help for people buying the "12-48 Volt 1800/2500 Watt ZVS induction Heater"
[Electronic circuits]
Bert911
June 24, 2019, 04:58:39 PM
post Re: Static electricity measurement
[Static electricity]
Doros
June 24, 2019, 01:56:17 PM
post Re: "LightningStorm" VTTC project
[Vacuum tube Tesla coils]
Avenger
June 24, 2019, 01:28:59 AM
post Re: Welcome new members, come say hello and tell a little about yourself :)
[General chatting]
Patrick
June 24, 2019, 12:47:16 AM
post SSTC interruptor not working
[Solid state Tesla coils]
juuzpo
June 23, 2019, 10:58:48 PM
post Re: Is one of these good?
[Dual Resonant Solid State Tesla coils]
Golddustpeak
June 23, 2019, 07:22:47 PM
post Re: Welcome new members, come say hello and tell a little about yourself :)
[General chatting]
Jeff B
June 23, 2019, 05:52:56 PM
post Re: Onetesla TS driver board for OWN coil need some help
[Dual Resonant Solid State Tesla coils]
dexter
June 23, 2019, 03:15:53 PM
post Onetesla TS driver board for OWN coil need some help
[Dual Resonant Solid State Tesla coils]
donnersm
June 23, 2019, 10:09:40 AM
post Re: Welcome new members, come say hello and tell a little about yourself :)
[General chatting]
Patrick
June 23, 2019, 10:07:04 AM
post Vacuum Epoxy Potting Attempt in Progress... (Pic Update 6/23/19)
[Transformer (ferrite core)]
Patrick
June 23, 2019, 08:50:14 AM
post Re: Si-wafers, ideas?
[Electronic circuits]
johnf
June 23, 2019, 01:28:20 AM
post Re: Welcome new members, come say hello and tell a little about yourself :)
[General chatting]
Twospoons
June 23, 2019, 01:17:50 AM
post Re: Si-wafers, ideas?
[Electronic circuits]
Twospoons
June 23, 2019, 01:04:13 AM
post Re: What happens when L and C are not matched?
[Electronic circuits]
Twospoons
June 23, 2019, 01:02:55 AM
post Re: Welcome new members, come say hello and tell a little about yourself :)
[General chatting]
Coupling
June 22, 2019, 11:10:02 PM
post Re: ZVS Mazzilli circ with active gate driver - what causes ringing?
[Electronic circuits]
flyrod
June 22, 2019, 08:45:17 PM
post Re: Help for people buying the "12-48 Volt 1800/2500 Watt ZVS induction Heater"
[Electronic circuits]
flyrod
June 22, 2019, 08:36:34 PM
post Impressive 500,000 watt radio station from the 1930's
[Electronic circuits]
petespaco
June 22, 2019, 06:18:22 PM
post Re: Help for people buying the "12-48 Volt 1800/2500 Watt ZVS induction Heater"
[Electronic circuits]
hightemp1
June 22, 2019, 05:47:22 PM
post Re: Help for people buying the "12-48 Volt 1800/2500 Watt ZVS induction Heater"
[Electronic circuits]
T3sl4co1l
June 22, 2019, 05:21:58 PM
post Re: Help for people buying the "12-48 Volt 1800/2500 Watt ZVS induction Heater"
[Electronic circuits]
hightemp1
June 22, 2019, 06:10:01 AM
post Re: Siemens Polyphos 30 xray machine
[X-ray]
Mads Barnkob
June 21, 2019, 09:08:16 AM
post Re: Choke-input DC-filter
[Electronic circuits]
johnf
June 20, 2019, 08:58:07 PM
post Re: CRT hacking ideas....
[Electronic circuits]
badpeter
June 20, 2019, 07:28:00 PM
post Re: Ideas needed!
[General chatting]
MRMILSTAR
June 20, 2019, 05:13:57 PM
post Re: Why did 4HV die??
[General chatting]
MRMILSTAR
June 20, 2019, 05:07:41 PM
post Re: Help for people buying the "12-48 Volt 1800/2500 Watt ZVS induction Heater"
[Electronic circuits]
Bert911
June 20, 2019, 02:14:55 PM
post Re: Choke-input DC-filter
[Electronic circuits]
kamelryttarn
June 20, 2019, 12:53:58 PM
post Re: Choke-input DC-filter
[Electronic circuits]
johnf
June 20, 2019, 12:09:27 PM
post Choke-input DC-filter
[Electronic circuits]
kamelryttarn
June 20, 2019, 11:43:58 AM
post Re: Welcome new members, come say hello and tell a little about yourself :)
[General chatting]
shrad
June 20, 2019, 08:57:42 AM
post Re: adventures in homemade flyback bobbins.
[Transformer (ferrite core)]
Patrick
June 20, 2019, 04:06:00 AM
post Re: Welcome new members, come say hello and tell a little about yourself :)
[General chatting]
johnf
June 20, 2019, 03:50:56 AM
post Re: CRT hacking ideas....
[Electronic circuits]
petespaco
June 19, 2019, 11:29:28 PM
post Re: A dynamical arc model v2
[Dual Resonant Solid State Tesla coils]
Uspring
June 19, 2019, 01:40:16 PM
post Re: Welcome new members, come say hello and tell a little about yourself :)
[General chatting]
Mads Barnkob
June 19, 2019, 10:49:45 AM
post Re: A dynamical arc model v2
[Dual Resonant Solid State Tesla coils]
Mads Barnkob
June 19, 2019, 10:43:32 AM
post Re: Light, laser and optics - Link collection
[Light, lasers and optics]
shrad
June 19, 2019, 10:03:54 AM
post Re: How go get single mode optic fiber to work
[Light, lasers and optics]
shrad
June 19, 2019, 10:02:21 AM
post Re: Welcome new members, come say hello and tell a little about yourself :)
[General chatting]
shrad
June 19, 2019, 09:35:52 AM
post Re: CRT hacking ideas....
[Electronic circuits]
badpeter
June 18, 2019, 11:35:16 PM
post Re: CRT hacking ideas....
[Electronic circuits]
Mads Barnkob
June 18, 2019, 08:22:23 PM

Sitemap 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30